[evlatests] low-band noise contamination update

[Paul Demorest]

hi all,

This is an update on the issue of the low-band noise diode signal 
leaking into L/S band.  About two weeks ago, the low-band system on ea09 
was modified to use a different noise source part.  Wes can provide more 
details, but my understanding is that the key feature is the new diode 
generates 20 dB less signal power than the older one.  Some attenuation 
downstream was removed so that the final in-band switched power signal 
level should be about the same.

I took data last week (6/14 and 6/15) to see if the leakage effects 
improved following this fix.  It turns out they do look significantly 
better!  The first two attached plots show the "on-off" switched power 
vs frequency spectra pre- and post- modification.  Sorry about lack of 
axis labels... I think the most useful plot is the bottom panel which 
gives the ratio of the switched power seen at S-band when the low-band 
cal is switching vs when the low-band cal is off.  The colors are the 
different polarizations.  Ideally we want to see a flat line with a 
value of 1.0.  With the old part the low-band leakage was up to ~4x as 
strong as the S-band cal signal in certain parts of the band.  With the 
new part this initial test couldn't pick up any effect at all.

I then ran a longer test and it turns out the effect is still present 
but at a much reduced amplitude.  These data are shown in the third 
attached plot.  The leakage is now at most ~4% of the S-band cal power.  
So this seems completely consistent with a 20 dB reduction in the 
leakage signal level.  A topic for future discussion is whether this 
level of effect would be acceptable or not (in case VLITE wanted to make 
use of the switched power for example).

I did not take any 4/P-band data to check that the in-band switched 
power measurement still looks OK with the new part.  Maybe Frank or 
others can comment on this aspect.

Cheers,
Paul
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ea09.olddiode.png
Type: image/png
Size: 128798 bytes
Desc: not available
URL: <http://listmgr.nrao.edu/pipermail/evlatests/attachments/20170622/78c8ce18/attachment-0003.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ea09.newdiode.png
Type: image/png
Size: 95176 bytes
Desc: not available
URL: <http://listmgr.nrao.edu/pipermail/evlatests/attachments/20170622/78c8ce18/attachment-0004.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ea09.newdiode.zoom.png
Type: image/png
Size: 160276 bytes
Desc: not available
URL: <http://listmgr.nrao.edu/pipermail/evlatests/attachments/20170622/78c8ce18/attachment-0005.png>